Speech Tech Market Entering Critical Period
- By John K. Waters
- March 1, 2005
The next 18 months will be a make-or-break period for vendors in the speech technology market, say the authors of a new report. For speech to become an effective enabling technology, one that enterprises take seriously, vendors must do more than simply automate touchtone activities, the report argues. They must deliver higher-order business benefits instead of incremental improvements, and they must develop a horizontal technology layer, an abstraction layer that allows the simple interfacing of speech with enterprise IT infrastructure.
The report from industry analysts at New York-based The 451 Group ("Speech technology market: Hanging up on growth?") focuses on automated speech recognition technologies (ASR), which allow users to access information and services via spoken commands.
The ASR market segment represents approximately 5 percent of the overall market for interactive voice response (IVR) technology and services; the other roughly 95 percent of the overall market is made up of IVR technology based on much older and simpler touchtone menus that direct users to select options by pushing buttons on a telephone keypad.
The report describes the speech technology market as "structurally messy," a state that has inhibited the creation of a coherent market taxonomy, and consequently, the normalization of discrete product categories.
"Vendors straddle multiple niches, and the market is characterized by the tremendous amount of labor-intensive custom work that vendors carry out for individual customers," the report states. "Even those companies and subsectors that appear on the surface as if they should be selling 'off-the-shelf solutions' are often providing highly customized (and subsequently high-cost) implementations that have questionable returns on investment for customers."
"Less-than-stellar customer-side return on investment from speech implementations stands in the way of the speech technology market making substantial breakthroughs in its business," says Steve Coplan, sector head for networks and media at 451, who wrote the report with support from Chris Noble and Steve Wallage, both 451 directors of research. "The total cost of ownership of speech applications is too high relative to alternative investments that are competing for corporate budgets. The amount of labor-intensive customization involved in building these applications compounds the issue. Services continue to account for a disproportionate percentage of total implementation costs, which leads us to believe that TCO estimates may prove unrealistic."
Speech tech industry watchers can expect to see competition heating up among vendors in the coming months, as well as an uptick in mergers and acquisitions, the report predicts. IBM and Microsoft are likely to emerge as the winners in this market, because of their strategic commitments to the sector, and their considerable resources. "But as it stands, 451 Group analysts don't believe either vendor has the tools or commitment to transform the technology and remove the complexity from the development of speech applications," the firm said in a statement. "Both IBM and Microsoft will need to buy that expertise, and speech market participants need to find a home in the broader enterprise IT firmament."
The 451 Group report recommends that enterprise software vendors and systems integrators that have held back from dabbling in speech consider the "significant potential of leveraging their investments in customer-facing Web infrastructure, business intelligence software and customer relationships to replace humans with systems in labor-intensive call-center environments."
The report was issued just prior to the start of SpeechTEK West, the speech technology industry's biannual industry gathering, in San Francisco (February 21-23).
For information, go to: www.the451group.com.